[Fwd: [Lowfer] BPSK and PSK31]

Andre' Kesteloot akestelo@bellatlantic.net
Fri, 23 Apr 1999 16:30:48 -0400

Lyle Koehler wrote:

> I'd also like to see a slower version of PSK31 for use on LF and in
> other weak-signal applications. Technically it wouldn't be PSK31 any
> more, but maybe PSK125?? The phase fluctuations on a typical HF
> propagation path actually make higher speeds perform better, but on LF
> the phase variations occur on a much longer time scale. I don't see any
> reason that the algorithms used for the sound-card version of PSK31
> can't be modified for slower baud rates, except that most of the
> interest driving the development has been for real-time
> keyboard-to-keyboard communication. Michael, DL6IAK already offers
> software for use with the SHARC EZ-Lite DSP board that adds variable
> transmission speeds (PSK16-124 ) with stronger QPSK modes. It's only a
> matter of time until some dedicated weak-signal experimenter (with a lot
> more programming knowledge than I have) comes up with similar features
> in a sound-card version.
> Lyle, K0LR
> John Matz wrote:
> >
> > Hi all
> > I would like to make some theoretical analysis comments on BPSK and PSK31.
> >
> > CW at 12 WPM is 1 character/second (roughly 10 baud).
> > BPSK as used by Lowfers is also about 1 character/second (about 10 baud
> > ASCII).
> > PSK31 is at about 3 characters/second Varicode (about 31 baud).
> >
> > This means PSK31 needs about 3 times the RX bandwidth, and roughly 5 dB
> > worse sensitivity than BPSK and 2 dB worse than 12 wpm OOK CW.  That's due
> > to the faster data rate in PSK31.  Also there is no way to run at a slower
> > speed or data rate and stay in PSK31, right?  In BPSK the data rate is
> > often slowed down to get better sensitivity, until the equipment or path
> > can't stand it.  The higher speed of PSK31 improves performance at higher
> > frequencies where the TX and RX phase noise and stability (and the tuning
> > and the path) aren't as good.  As people have observed, a bit faster data
> > rate than 10 baud seems to be better at higher freqs.
> To unsubscribe, send to MAJORDOMO@qth.net "unsubscribe lowfer" (Do not
> send to list!!) Send on list submissions to lowfer@qth.net